Skip to main content

Full text of "The hiding power of white pigments and paints / by A.H. Pfund."

See other formats


JBK 2 

The Hiding Power 

of White Pigments 

and Paints 




Issued by 

The New Jersey Zinc Company 

(Established 1&18) 

160 Front Street, New York 

The Hiding Power 

of White Pigments 

and Paints 

A. H. PFUND, Ph. D. 

Associate Professor oi I'l.. 
Johns Hopkin I 

With a Discussion on Vpplications of the 

Hiding Power-Brightness Relationship in 
Comparing White Pigments and Paints 

B H \ NELSON M V and G I V. STUTZ, Ch.J 
Paint Section, Research Division 
The New Jersey Zinc Company 



l,.K L926, Revised 

The New Jersey Zinc Company 

(Established lHlHi 

160 Front Street, New York 

Pnnlu. I- I ti-.lril.ulri) by 

The New Jersey Zinc Sales Company 

[no irpoi i 
New York Chicago Pittsburgh Cleveland San Fi sco 

/ - 

The Hiding Power of White 
Pigments and Paints 

By A. H. Pfund, Ph.D. 

v . i lit- Professor <>l Physics, Johns Hopkins I niv< 

Tl 1 1 : hiding pnw«i of a paint ma\ be defined as that property of a 
paint which enables it to obliterate beyond recognition an> back- 
ground upon which it may be spread, hi order to compare the 
hiding powei of two white pigments, the almost universal custom is to 
rub down equal masses of the two sample, with the same amount of linseed 
oil and ultramarine blue. The resultant pastes arc spread side b> side and 
the sample which has the paler tint is supposed to contain the pigment 
of greater biding power. The idea is that the particles of white pigmenl 
hide the dark particles and hence give a light tint to the past,-. \^ a 
result of the experiments to be described presently, it appears that the 
criterion, upon which the above tests are based. U incorrect. 

The object of this investigation is to present a method which will 
yield numerical values of the true hiding power of white pigments and 
paints. The basic idea underlying the discussion is this: granting that 
an infinitely thick layer of a paint will "hide" a given background com- 
pletely, it is sought to find the thinnest layer which will hide the back- 
ground as effectively as does the infinitely thick layer. Obviously, the 
thinner the layer of paint required, the greater is the hiding power of the 
paint, tor white paints, the severest tesl met with in practice is a white 
wall with black lettering. It is desired to obliterate the lettering by cover- 
ing the entire surface with successive coats of paint. These conditions 
are simulated in the instrument about to be described. 

The form given the instrument is shown in Fig. 1. Here, \ is a plate 
of glass, 14 x 5 x .6 cm., whose upper surface is optically flat. The lower 
surface is coated with black baking enamel, which > \M> the desired black 
background. A transverse groove, B, about 2 mm. deep and 1 cm. wide, 
is cut in the upper surface and a millimetre scale is etched, as shown in 
the drawing. Resting upon plate \ is plate C (7 x 3.5 x .6 cm. whose 
lowei >urface is likewise optically flat. A strip of thin steel, D, 0.45 mm. 
thick, is attached to C, so that a wedge-shaped layer of white paint ma> 
be formed between the plates. This wedge terminates abrupth at the 
• thick" layer, 15. and. so long as the hiding is not complete, the 
line of demarcation is visible. Ii\ sliding the wedge to the left it is finally 

I in \i w ii K-i ^ /inc COMPAN1 

impossible to see fli< ed I m a knov I le ol the w( dg< 

and the de, ii is possible to calculate thi Lhickness «•! 

this critical layer lying immediatel) above theedgi B Vo^ in advani inf 

the plate C until tin line of demarcati :an no longei b een we have 

done it, so to speak. I rrect this, we must reversi th( motion ol 

the wedge until thi n ju i I" distinguished ovei its entin length. 

I .1 to disap] 
ind appearance ol il d 1 1 n - desired result Sina thi 




.'.',40 J 




, ,- 

I i K I 

lue I the fact t hat i hi l< u i 
• hi< L the human eyi i an detect i 

i i I 

Ui I pel 

1 .l.l< 

In. I.I II I |,. I "III. 

I Ii. 


Mil IUIHN4. row M( OF will IK. PIGMENTS v\n ni\h 

of the infinitely thick layer, then, h\ means of a photo-electric cell, which 
measures the light from the two layers in contad a1 the edge B Fig. I 
it will be possible to locate the desired poinl vers accurately; 2) b\ mak- 
ing, preliminarily, a stud} of the variation of brightness of the painl with 
increasing thickness, a formula ma> be established connecting these two 
quantity-: then, by finding a position ot" the wedge where the brightness 
is, sa} one-half that of the infinitely thick layer, it is possible to calculate 

t,_ j 

the thickness yielding 99 per cent, of the brightness of the infinitely thick 
layer. These methods have not been developed for the reason that the 
cryptometer in its present form is not only sufficiently accurate for most 
purposes, but is extremely simple in it- construction and operation. 

A sharp distinction must be made between the hiding power of a 
pigment and that of a paint. Not only are these quantities expressed in 
different units, but they are not necessarih related in the sense that a 
pigment of great hiding power necessarily produces a paint of correspond- 
ingly great hiding power. Taking up first the hiding power of pigments, 
let us consider an intimate mixture of x grs. of a white pigment and y grs. 
of colorless (or very pale) linseed oil. This mixture is tested in the crypto- 
meter and the critical thickness producing complete hiding is found. 


THE \KW JERS1 ^i /Wt ( OMPAN1 

t = thickness of critical la>cr (in cms.). 

b = numbers of grs. of pigment in a disc of 1 cm. 2 base and thickne^ I 

Then, if b grs. pigment hide I sq. cm., we find the number of sq. cm. \ 
covcn-d and hidden b\ 1 gr. of pigment from the relation 

I) : i I \ or = A. 


e the hiding power Is better the thinner the layer, i.e., smallei than 
b. we ma> define the hiding power of a pigment a- (he re< ipn* .,1 of the 
number of grs. of pigment, mixed with colorless linseed oU to painting 
consistency, which are nea bid< i black. aon-absorben1 area 

I cm 2 . This is numerically equal to the Dumber of square centimetres 
covered and hidden b> I gr. of pigment. Hiding powers of pigments will, 
therefore, be expressed in terms of cm. 2 pei gi Experimenl has shown 
that the hiding power of th. pigm( ul is affei ted b} the relative amounl 
of oil present the thinner the mixture, the greater the hiding powei 
This means that the hiding pow< i,K a function of the aumbei 

of white particles, but also of their separation. In ordei to reduce all 
mixtures to a standard condition, the terms ''painting consistency" .in 
specified in the above definition. I >ata will be presented latei 

Turning next to the hiding p paints, we chooa as a measure 

the number of square feel which one gallon of paint will covei and bide 
While the retention of the i I st m might at first thought -• m 

advisable the selection of the English systi m oi units. 

1 ' actual m< ind calculations are ver$ simple I be paint 

is measured up in the , iwledge oi the critical 

thickness the uumbei of square i gallon is calculated at ona 

Xs fi Illil,trr pi mad< so that on< maj r< ad 

tne result directly , i, bed on th lowei 

n "* rayta istic pigment* and paints are 

summarized in th«' follow ing table. 


I Sublimed w bite l< 

i I • • • 

/ , . | . ■< | 









tiding powei 


1 Vim. hi 


sq ft 



24 1 





mi miiiM. ru^itui whim I' m- \m> i'UM- 

Concerning the pigments themselves, the results obtained speak for 
themselves The theory of hiding power as dependent on particle size, 
wave length ol light, character of vehicle, etc., will be taken up in a sepa- 
ra te paper. The one point, however, to which attention musl be called, 
is that involving a comparison of pigments 6 and 7. The <>nl\ difference 
between these two pigments is thai n is white while 7 is pearl gray. If 
these same pigments are compared l>> rubbing them down with oil and 
ultramarine blue, the gra} sample will show the darker tint; hence it will 
Im r..i^id.-ir ( | as having a smaller hiding power than the white sample. 
As a matter of fact, the hiding power of the gi -.-. unple is more than thre< 
times as greal as that of the white, [n a paper describing a new colorimeter,^ 
il will be shown thai all paints are more ... less gray. I» is, therel 
deal that, since the rundown tesl yields results which are violently a1 

. iri, with those obtained with the cryptometer, the former method 

musl be discarded 

The results on paints are interesting. While the hiding powers 
of zinc pigments are largei than those of the whit. l< ids the reverse is 
true of the painl I ompar< samples 2 and 6.) The reason is obvious 
w hen the "oil absorption" of the pigments is considered. Our pound of 
zinc oxide mixed with one pound of linseed oil yields a mixture of proper 
painting consistency, while a similar mixture of white lead in oil is entirely 
too "runny" for painting purpose- \I«m lead must be added t<> give 
the coned "body" or thickening to the paint. Such a painl has the 
_,,,,!.. | lh |.i,_ |...wn but it hit- irained its superiority in consequence oi 
thr ,An-> ..t I- id pigment which lias been added. 

While ideal conditions exisl in the cryptometer tests, it seei 

of interesl t mpare the results thus found with those obtained in 

.,, tual painting practice. I am indebted to Mr. R. J. Hank for preparing 
the paints and carrying out the painting tests. Briefly, the experimenl 
consisted in preparing three zinc paints pigmenl ground in linseed oil 
II,,,, clapboard panels, having an urea of 3.75 square feet, were well 
primed with white paint. \ cross of black paint was applied t 
these surfaces and the whole was allowed to dry. Then successive i 
of the respective paints were applied until the bla< k i ross i ould ao I 
be seen From a knowledge of the area painted and th< irolume ol paint 
consumed in ea< h i ase il was possible to calculate the number of sq 
feel per gallon which the respective paints would covei ind hide I 
tests extended over a week oi more When aboul half the adeq 
Dumbei of coats had been applied, the writer made hiding b sts 

with the i t sptometei The r.-tilN «.■!»• .1- l-ll-w- 

— ^ . Baltimore 1918 also at meeting 

ol the * >pti« 'i S tj B iltimon Dea mb< 1 

iiik Nhw imsn /i\< i iimpwv 

No. of Hiding-power Hiding-power 

Paint. Coats Painting Test. Crvptometer. 

sq. ft. ' sq. ft. 

ffal. gal. 

1. Zn< > i heavily leaded) 63.2 per cent. ) 6 225 256 

Vehicle 36.8 per cent. J 

2. ZnO (less heavily leaded ). 52.7 [hi . .rit. 1 7 204 213 
Vehicle 17.3 per cent. / 

3. ZnO (pure) 1-8.6 per cent. '» 150 150 
Vehicle 51.4 per cent. / 

On the whole, the a-reernenl is quite ^ood. The findings of the 
painting test will always be the smaller for two reasons: 

1. Brush mark- will leave thin places which must be covered up. 

2. <)nl\ whole numbers of coats may be applied, i. e., if theoretically 
6.4 coats would just hide the black cross it is unavoidable that 7 coats 
be applied. Obviously, the greater the number of coats, the closer is 
the approach to the ideal conditions which are realized in the cr> ptorneter. 

The conclusion to be drawn from these experiments is that the true 
hiding power of a paint may he obtained in a very few minutes by means 
of the cryptometer. 

M is clear that if a surface of lighter tint (say, white pine) is to be 
hidden, the number of square feet per gallon will be considerably larger. 
The cryptometer musl necessarily leave out of consideration the absorp- 
tion of the paint in the pores of the wood. However, this instrument 
ma} be adapted to take care of lighter backgrounds by removing the black 
paint from the lower surface of plate (A), Fig. 1. If this plate be rested 
on the actual surface to be painted, it is a-ain possible to determine in 
advance just how main gallons are required to paint a -riven area. 

The principal purpose of this discussion is to presenl the crypto- 
meter rather than final values of the hiding power of pigment and paints. 
In order to realize the latter, it will be necessary to -rind the pigments 
in oil and to formulate a rigorous definition of, and tesl for, "painting 
consistency ." 

The results obtained rna> be summarized as follows: 

1. An instrument whirl, yields numerical values of the hiding powers 
oi pigments and paints has been de> ised 

2. \ table of true hiding powers of characteristic pigments and 

p.iinls iv presented. 

I \ comparison of actual painting and cr> ptorneter tests shows 
that the two are essentially in agreement. 

This tfork was carried oul parti} at the Johns Hopkins I oiversit) 
and partly at the Research Laboratory of The \<w Jersey Zinc ( 

h "i"" '•' , '""" the ' Journal of it..- Franklin [nirtitiili 




Application of Hiding Power — Brightness 

Relationship in Comparing White 

Pigments and Paints 

1^ n. \ An -•• - M V vm> <i. F. A. Sttjtz, Ch. E. 

SINCE the publication of the preceding Research Bulletin on this 
subject, furthei investigation i has established a useful relationship 

between the brightness and biding power characteristics of white 
pigments. I In- work indicates that within the range of the brightnesses of 
the eommereialh produced white pigments, which is from 
brightness, the brightness-hiding power relationship is prartieall\ linear. 
anc ] thai this line, it projei ted to the origin, cuts the axes very nearl> at the 
theoretieal point. 100',' brightnev* hiding power.* 

The practical advantage of this, as pointed out in the original investiga- 
tion, is the ease with which n.rnmereial white pigments and their paint 
combinations) may be compared on a basis of equal brightness. It is, then, 
only necessary to make observations for both brightness and hiding power 
on the samples in qu.-t ion II these are plotted, and a line projected from 
this point to the region of the point 100',' brightness —0 hiding power, 
comparisons may be made at desired brightnesses. 


Fig. L, showing this relationship on a l»,i>h «»!" hiding power in sq. cm. 
per gram of pigment, is taken fr< an the original paper. The data are for the 
pigments alone ground with pale linseed oil without turpentine or dryer 
added. The circled dots are the readings as taken on the original pig- 
ments. The crosses are the readings obtained by Pfund when the bright 
nesses were cut down with bone black, and illustrate the nature of the 
brightness -hiding power relationship referred to above. 

Reference to the following figures and tables will show at a glance 
what erroneous conclusions as to the relative merits of pigments may be 
arrived at by confining the observations to hiding powers alone. 

» A. H. Pfund, Journal of The Franklin Institute, Julj 1923- 

tThr i. ,.|. risn ferred to the original paper for a more detailed dia ussu i the ezai t lorn, for uua 

I UK \hw JKHM > /IN* < o\1l'\\\ 


Per ce/?f Br /$/?/■/? ess 
Fig. 1 

Hiding power-brightness relationship for white paints, as determined 
l.\ Pfund. (Journal of the Franklin Institute, July, 1923). 

In Table I and Figure 2 are shown the hiding power-brightness values 
for the most recent samples of the several white pigments obtainable in 
March 1926. In each case (he pigments were rubbed in pale (almost 
coloi less linseed oil. 

The higher brightnesses given in Figure 2 as compared with Figure 1 
ma} be due to three factors, first, improvements in the manufacture of 
the pigments used, second, difference in the oil used in making the pastes, 
and third, differences in the calibration of the instruments used. 

Two standard- I mi , !0 m pa ri son appear in Table I. They are: I the 
inal samples as obsei ved; 2) at a calculated brightness of 80%, which 
iffii i< ntl> low to be easil> attained by an> clean pigment which is suf- 
fi< ientl} white to be classed as a white pigment. 


I he same relationships may be used in comparing paints at painting 
consistency. It should be noted, however, that because of the var) 
amounts of vehicle required with differenl pigments to attain painting 


To 280 Sf. Cm per 6 ram 
at 807.Br/$/?f/7e$s. 

To 33/ SfCm per Gram 
at 80% Br/g/?tr?ess. 

Per cent Br/$t?tr?ess 

Fig. 2 
Hiding power-brightness values of white pigments as determined in a 

vehicle consisting of pale refined linseed oil. (See Table I). 




o £ 

w w as p 



+ + + 

: ^ + + + 

*l o\ ?l 

+ + + 

w — i- ^r i~ i- o> — 
+ + + + + + + + 


% I °° | 2 

T 5.— - *" - = 

— CI M 


~ - 7 


~ 'J, 

- - ■* z 






— "■ 

/ / - \ \ 



I UK llllllM. IMIftKK OF Will IK I'K.MIM^ \M» I'VIMS 

consistency, the hiding powers -»l the paints may or may not hold the 
same relative order of magnitude as for the pigments themselves. 

A number of representative paints purchased in the open market, 
and others prepared in the laboratory, were compared for hiding power 
and brightness. Hiding powers were measured on an improved crypto- 
meter which will be described in another place (see Fig. 6), 

Brightness measurements were made with the Pfund colorimeter 3 on 
the wet paints as applied on a steel colorimeter plate. The readings are for 
green light {A ^°)- which gives essentially the same results as white 
light, and enables the operator to take readings more easily and accurately. 
The maximum variation observed for the cryptometer readings was 
+ 6%. Generally the readings checked within + 3%. A maximum 
variation of + 5% is the limit observed for a large number of readings 
taken in the course of routine examinations of paints by this Laboratory. 
\\erages for brightness measurements, by two operators, checked 
within +0.2', excepl in one case where the variation was Y 0.4%. 

The compositions of the paints in question, together with the observed 
value> I'm hiding power and brightness, and the comparative hiding powers 
at a standard brightness (80%) are summarized in Tables II III and I \ 

The values for hiding powers at brightnesses other than the observed 
brightnesses for the samples, were obtained b\ extrapolation along the 
lines drawn from the observed point to the region loo\ brightness 
hiding power, in the manner previously explained. These are plotted 
in Figs 3, I and 5. The selection of a -ingle standard of brightness at 
which comparisons are to be made would be desirable. The somewhat 
different brightness ranges of the different types of paints also suggests 
that a standard mighl be selected for each type. For example, 85^ ap- 
parently is a relatively high brightness for an outside paint but a low 
standard for enamels, and also comparative^ l<>w for high grade flat whites. 
However, in the original report it has been indicated that it is not 
entirely practicable to extrapolate toward higher brightnesses be< 
certain pigments ma> already be essentially "clean" at the bright • 
observed. In order then to obtain higher brightnesses the physical 
properties of tic pigment it-It particle size, etc.), must be changed. 
This would give rise to what is essentially a ne* pigmenl with a bright™ 
hiding power curve of its own. In ordei to overcome this possible objec- 
tion, a standard brightness has been selected for these comparisons which 
is so low that all pigments classed as white ma> be expected to attain the 
standard. The value thus suggested as a standard for comparison is 
80% brigh tness. 

•Pfund, "The < olorimeto «•' Nearly White Surfaces I Frank. Ins... March, 1920. 




Bo so 

Perce/pf Bn$/?//7ess 


i I ,1,1. II 


nwalnatureol theresull , „, ,, ,.,,,,,. 

which are shown, cert 


Oulside White Paints 

1 mercial tidi »l..i. 

ind I. ,1,1. II [N ,„ ,,,. 

"" Mi ' 11 ' unberoj 


.1. from 



— ~ 3 r- CT I - i - sC c i- Tl 

r- "I Tl Tl Tl Tl Tl Tl Tl 


!t L N 

= - _ 

w tt 'l ~ < c ' ?l N ' ri Cl Cl Cl N 



■- »13 

Tl I - I - 

» C - 

-^- ;> 









._ z 


5 ; : = 

= =: s 

— — u 

1£ — 


i- :: g 


TIIK NKW JHt-n /Wi (d\l|'\N^ 


Per cer?r Br/^/?r/?ess 

Hiding power-bri.i ds and gloi* mill white 

I ibl. I\ 




5 (»S lc 

j- ^"^c: « -" m - « — 
3 •— i * ' - — 
U ~= - 7 

E r " w ^ l 2 

r - % = 

> s 







— Tl C I T I 

x£ - - 

u u 

lO irt '" 


Nil Mtt JIHH \ /J\< COMPAN1 

Flat \\ hile Paints 

tsofthifi i more immediate opportunity foi the develop 

men! ol highei standards o( brightness and obscuring power, because thi 

ments \n hi< h can I" i not limit •<! I nsideral ion "I 

pigments .i\ ailable 1<>« use in ext( 

from ihe beaut) ..I .( clean white surfacj where white is d< 

i.l. iid o( bi i-lii ness I hi 
i ioi decoration, i an onl) l>< obtained 

i< imp* w i. 

\Mi- ■ I d< pr< i iating th< 

i]ualit> ol tin painting job. it In»..i \<> I n 

1 1 1 V 

Lion both bj thi 

i : ill n prod 






Fig. u 

Iniprmnl Type of ( "rvptorneter 

The usual metallic strip attached to the top plate i^ replaced l>\ two 
small metallic squares at A and \ : again, the ^lass in the central area at 
C is cut away for a distance of 5 mm. When this plate is laid upon the 
lower one, contact is made* only on the outer, narrow strips at \ \ and BB . 
I pun tilling in the end> of the liau-\riv ( . -,oo\r. I), it i^ found that, when 
paint is applied to the area between the longitudinal grooves, the upper 
plate may be moved back and forth many times without dragging paint 
to the outer strips alon^ which the plates are in contact. Needless to sa> . 
no pressure othei than the weight of the top plate is now necessary . Ver} 
consistent results are obtained as a result of these modifications. 


New Jersey" 

The World's Standard for Zinc Products